President Lee Jae-myung has ordered a complete overhaul of administrative procedures to expedite the “3 Major Mega Projects” announced by South Korea’s leading companies, including Samsung and SK, a combined investment totaling 4,700 trillion won (approximately $3.1 trillion). He declared a strong push for speed, mandating that licensing processes, which previously took up to six years, be conducted simultaneously to eliminate wasted time, and that speculative land purchases—a persistent obstacle to large-scale industrial complex development—be blocked.
At a meeting held at the presidential office on the 6th, President Lee emphasized, “From now on, procedures that were previously followed step-by-step, from site confirmation to groundbreaking, should be carried out concurrently.” Specifically, he ordered that negotiation and compulsory procedures in the land expropriation process be pursued in parallel. He also directly refuted political opposition, stating, “It doesn’t add up to call it a fraud or impossible while simultaneously asking why we were left out. I hope they will not obstruct the process.”
The President’s directive is a measure to support the 4,755 trillion won (approximately $3.1 trillion) private investment plan announced on June 29 during the “Korea’s Great Leap Forward: 3 Major Mega Projects National Report” at the presidential office. Samsung Group will invest 2,655 trillion won (approximately $1.7 trillion) in a semiconductor cluster in South Korea’s Honam region, a High Bandwidth Memory (HBM) packaging fab in the South Chungcheong Province area, and a humanoid robot and smart manufacturing hub in the Yeongnam region. SK Group will also pour 2,100 trillion won (approximately $1.4 trillion) into an AI memory production belt connecting Yongin, Cheongju, and the Honam region, along with AI data centers and its semiconductor business. The combined investment of the two companies is the largest ever, amounting to 6.5 times this year’s government budget (approximately 728 trillion won).
Investment Tailwinds Spread to Semiconductors, Robotics, and Construction
The securities industry views this mega project as an opportunity to ease the “K-shaped market” concentrated in Samsung Electronics and SK Hynix, diversifying the investment axis. Park Sun-young, an analyst at Mirae Asset Securities, analyzed, “Investors need to prioritize AI robot finished products, core components, and physical AI development companies that are directly linked to government strategy.”
Indeed, immediately after the project announcement, the stock market saw construction companies based in the Honam region, such as Kumho Engineering & Construction and Namhwa Construction, hit their daily upper limits, while large construction stocks like Hyundai Engineering & Construction and GS Engineering & Construction also posted double-digit gains. This reflects expectations that demand for various infrastructure—including roads, water supply, power grids, and housing—will expand alongside the semiconductor plants.
Interest is also focusing on power infrastructure-related stocks. As AI data centers are facilities that consume massive amounts of electricity, investment in transformers, ultra-high-voltage cables, and power transmission and distribution equipment is essential. Consequently, the three major power equipment companies—LS Electric, Hyosung Heavy Industries, and HD Hyundai Electric—have been drawing investor attention since the government’s announcement.
Hyundai Motor’s moves are also noteworthy amid the government’s policy of targeting physical AI as a future growth engine. Starting next year, Hyundai Motor plans to invest a total of 9 trillion won (approximately $5.9 billion) in phases on a 1.12 million square meter site in Saemangeum to build an AI data center and a robot manufacturing and parts cluster. Samsung Electronics has also been accelerating the development of humanoid and industrial robots since acquiring Rainbow Robotics in 2024. At the national report, Samsung Electronics Chairman Lee Jae-yong stated, “Demand for robots will expand beyond factories to society as a whole, including industrial sites, homes, hospitals, and restaurants,” and announced that robot investment would be concentrated in Gumi, North Gyeongsang Province.
Micron Builds HBM Plant in Hiroshima, Japan… Subsidy Competition Intensifies
Meanwhile, global competitors are also moving quickly. Micron Technology has broken ground on a next-generation HBM production facility in Hiroshima, Japan. Of the total investment of 1.5 trillion yen (approximately $9.2 billion), one-third is being subsidized by the Japanese government. The assessment is that the decisive factor was Japan’s government-led structure, which defines semiconductors as an “economic security asset” and pushes them at the national level. The advantage of being able to source 80% of parts and equipment locally, thus lowering supply chain risk, also played a role.
For Samsung Electronics and SK Hynix, this means a competitor is increasing its production capacity (CAPA) backed by more generous state support. Global subsidy competition is expected to heat up further, coinciding with the launch of South Korea’s special semiconductor committee.
KT Transforms into a ‘Token Seller’… Enters AI Billing Market
The transformation in the telecommunications industry is also notable. KT has declared its shift from a telecom company to a “token seller,” making a full-fledged entry into the financial and AI billing market. With the traditional telecom business being stable but offering little prospect for significant growth, the strategy is to apply its telecom experience and capabilities to non-telecom sectors. In the past, KT attempted unconventional ventures like virtual goods but did not achieve major success. The industry is paying close attention to how this new challenge will unfold.
Domestic AI Must Prove It’s ‘Worth Paying For’
The results of the second evaluation for the government’s “Independent AI Foundation Model” project are expected soon. LG AI Research, SK Telecom, and Upstage have completed testing, and Motif Technologies will join by the end of this month. However, the industry’s reaction is lukewarm. Critics point out that providing confidence that a model is “worth using” in real-world applications is more important than benchmark scores. With global models like OpenAI and Anthropic, as well as free open-source models, already performing simple tasks excellently, the domestic AI industry has reached a stage where it must answer the question, “Why specifically use a domestic model?”
In the AI Era, ‘What You Built with AI’ Matters More Than ‘Years of Experience’
The advancement of AI is also fundamentally changing hiring market standards. According to a Digital Daily report covering 12 South Korean IT companies, hiring criteria are shifting from “whether you can use AI tools” to “whether you can define problems and solve them with AI.” A particularly noteworthy phenomenon is the “irrelevance of career years.” Developer A, a new hire at Buzzvil, single-handedly implemented an entire CRM system just three months after joining the company. This was a task that would typically be tight for a team of five-year veterans working for several months, but was accomplished solely through the ability to know precisely what to instruct the AI to do and how.
As a result, some companies are abandoning the hiring of developers who only code and are shifting their recruitment focus toward “Forward Deployed Engineer (FDE)”-type talent who can handle everything from problem definition to implementation alone. Going forward, “what you have built with AI” is highly likely to become a more important qualification on a resume than years of experience.